Proposal Abstract: Project Background: Multiple genetic markers for colorectal cancer (CRC) risk have been identified, and while practical tests for CRC risk are not yet available, such tests are both plausible and representative of the types of applications of genomic medicine currently being contemplated. Such tests could be used in the future by VA primary care clinicians to tailor cancer screening regimens to each individual's personal risk. Unfortunately, current approaches to communicating genetic risks (developed for genetic counselors to use with Mendelian diseases) are impractical and perhaps unnecessary in a VA primary care context. Furthermore, patient acceptance of genetic testing initiatives within VA may depend on the communication techniques used by VA clinicians to discuss test results, on whether the particular attributes and formats of the tests performed match patient's prior expectations about genetic tests, and on whether the reasons used to justify genetic testing make sense to veterans. Project Objectives: This proposal will experimentally evaluate different methods of communicating CRC-predictive genetic test results to VA patients in order to identify those approaches that (a) optimize comprehension and recall of the information, (b) generate consistent levels of risk perceptions and perceived vulnerability across different types of patients, (c) evoke the greatest trust in the validity of the test and (d) result in guideline-adherent CRC screening intentions. It will also seek to identify what properties of a CRC-predictive genetic test and justifications for genetic testing influence VA patient expectations about the test, their acceptance of the use of genetic testing for their care, and/or their reactions to the test results. Lastly, it will assess how well VA clinicians' beliefs about what matters to veterans in this context correspond with VA patient's actual beliefs and attitudes. Project Methods: The project consists of three distinct elements: (1) A series of 4 iterative surveys of veterans and family members (1000 per survey) at 2 VA medical centers will incorporate randomized controlled manipulations of hypothetical scenarios to identify which communication factors and test attributes most influence patient understanding and reactions to genetic test results. Examples of factors include presenting results as a list of mutations vs. an aggregate risk score, whether the DNA sample is derived from stool vs. blood, and whether the patient is being tested due to a family history of CRC or a broader age-based criterion. Results from each survey will be used to identify active factors and thus inform the design of later surveys. (2) A national mailed survey of 2000 veterans age 40 and older will confirm and expand our findings to veteran populations outside of Southeastern Michigan. (3) A national mailed survey of 500 VA primary care clinicians will assess whether clinicians understand what factors affect patients' reactions to genetic test results.